Day-of polling for last-hour panic-drop listings. Iterates a list of upcoming events and fires per-event when a steep discount appears within the final hours before tip-off.

Game-Day Fire Sale Watch

Day-of polling for last-hour panic-drop listings on the secondary ticket market. Iterates a list of upcoming events; fires per-event when a steep discount appears in the final hours before tip-off.

What it does

The monitor's for_each: walks a literal list of event IDs. Per event, on each fire:

The check calls a ticket-pricing endpoint for the cheapest 5 listings.
The when: arm tests TWO conditions simultaneously: - hoursToEvent <= 6 — only fire in the final-hours window. - discountPct >= 40 — only fire when the price is significantly below today's median for the event (i.e. a real fire-sale, not just a baseline cheap listing).
If both pass, a per-event notification fires with section/row/URL. Otherwise nothing fires.

This is the "opportunistic late buy" pattern. Outside the 6-hour window, the monitor stays quiet. Inside the window but with normal pricing, also quiet. The notification only fires when something genuinely unusual happens: a seller slashing to recoup anything rather than eat the ticket.

Data-source setup — read this first

The data-source story is identical to event-price-watch. You wire one scraper / API endpoint that returns the expected JSON shape; both monitors use it.

This monitor additionally requires the scraper to publish a discountPct field — the percentage discount of the cheapest current listing vs. the event's typical/median price (either today's intraday median, or the price ladder observed earlier this week). Most marketplace scrapers can compute this with one extra query.

Expected published bindings:

Binding	Expected value
`{{check.lowestPrice}}`	Cheapest current listing (number).
`{{check.eventName}}`	Display name of the event.
`{{check.section}}`	Section identifier of the cheapest listing.
`{{check.row}}`	Row within that section.
`{{check.hoursToEvent}}`	Hours from now to tip-off (number).
`{{check.discountPct}}`	Percent discount of `lowestPrice` vs. typical (number, e.g. `42` means 42% off).
`{{check.listingUrl}}`	Direct URL to the listing on the marketplace.

Edit before installing

In monitor.yaml:

for_each.items — replace with your event IDs. The monitor only fires inside the 6-hour window for each one, so listing dozens of events is fine; almost all of them will be inert most of the time.
for_each.as — loop variable name; default eventId. If you rename it, update the {{eventId}} reference in the check URL.
when[0].condition — adjust thresholds. hoursToEvent <= 6 and discountPct >= 40 are aggressive defaults; loosen to <= 12 and >= 25 for more sensitive alerting.

Schedule guidance:

Default cadence: every 5-10 minutes. Outside the 6-hour window the monitor is essentially a no-op, so high cadence is cheap.
For an even lower-noise version, schedule the monitor to fire ONLY during the relevant time window (e.g. 5pm-10pm if your events tip off at 7-9pm).

Bindings used per iteration

{{eventId}} — current event id from the list.
{{check.*}} — see "Data-source setup" above.

Tools used

http — core InTouch HTTP task type.

Pattern showcase

Combines literal for_each iteration + compound boolean condition. The condition uses && to require BOTH a time-window match AND a price threshold — neither alone is enough. This is the canonical "two-factor alert" shape that keeps notifications honest.

Related monitors

event-price-watch — far-in-advance hourly threshold check on ONE event.
good-seat-ai-watch — AI judges value across current listings.